Preictal connectivity dynamics: Exploring inflow and outflow in iEEG networks

Area of Science:

• Neuroscience
• Epileptology
• Computational Neuroscience

Background:

• Refractory epilepsy treatment relies on precise seizure onset zone identification via intracranial EEG.
• The epileptogenic zone concept has evolved to an epileptogenic network, highlighting interconnected brain regions.
• Understanding preictal network dynamics is crucial for optimizing epilepsy surgery outcomes.

Purpose of the Study:

• To investigate preictal network interactions between the seizure onset zone and the broader network using directed connectivity measures.
• To evaluate the effectiveness of connectivity metrics in identifying seizure onset zones.

Main Methods:

• Utilized a multicenter intracranial EEG dataset (243 seizures, 61 patients).
• Applied directed transfer function and partial directed coherence to analyze preictal connectivity.
• Computed inflow and outflow metrics for seizure onset zone electrodes.

Main Results:

• Observed two distinct preictal connectivity patterns: increased inflow to or prominent outflow from the seizure onset zone.
• Patients with grid/strip/depth implantation showed significantly higher performance (AUC) than those with ECoG or sEEG alone.
• Lesional cases showed a greater proportion of high-inflow and high-outflow patterns.

Conclusions:

• Seizure generation involves dynamic shifts in brain functional network information flow.
• Preictal connectivity patterns highlight network reorganization in epileptic transitions.
• Epilepsy is a network phenomenon, supporting network-based approaches for surgical localization.